Use of Near-Infrared Spectroscopy to Estimate Fiber and Crude Protein Content in Fodders
Objective: Demonstrate the need to use locally generated data in the calibration of a near-infrared spectrometer (NIRS) in order to predict the chemical characteristics of fodder; instead of using data bases from other geographic regions, as is commonly done in Mexico. Design/Methodology/Approach: Two groups of samples collected in prairies of the central highlands of Mexico, the first group was used to calibrate the equipment; the equations generated were validated with a second group, collected in prairies that were different from the ones of the calibration group, but in the same geographic zone. Results: The best regression coefficients of the NIRS predictions, compared to traditional laboratory analyses were for crude protein (CP), neutral detergent fiber (NDF), acid detergent fiber (ADF), acid detergent lignin (ADL), dry matter (DM) and organic matter (OM) (0.93, 0.87, 0.87, 0.56, 0.72 y 0.68 respectively). The lowest predictive value was observed in ashes (0.27). Limitations of the study/implications: The results show the need to use local materials in the calibration process. Conclusions: NIRS will make predictions of their chemical composition, since this is influenced by geographic origin of the sample and its botanical composition
| Main Authors: | , , , , , |
|---|---|
| Format: | Digital revista |
| Language: | spa |
| Published: |
Colegio de Postgraduados
2022
|
| Online Access: | https://revista-agroproductividad.org/index.php/agroproductividad/article/view/2231 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Objective: Demonstrate the need to use locally generated data in the calibration of a near-infrared spectrometer (NIRS) in order to predict the chemical characteristics of fodder; instead of using data bases from other geographic regions, as is commonly done in Mexico.
Design/Methodology/Approach: Two groups of samples collected in prairies of the central highlands of Mexico, the first group was used to calibrate the equipment; the equations generated were validated with a second group, collected in prairies that were different from the ones of the calibration group, but in the same geographic zone.
Results: The best regression coefficients of the NIRS predictions, compared to traditional laboratory analyses were for crude protein (CP), neutral detergent fiber (NDF), acid detergent fiber (ADF), acid detergent lignin (ADL), dry matter (DM) and organic matter (OM) (0.93, 0.87, 0.87, 0.56, 0.72 y 0.68 respectively). The lowest predictive value was observed in ashes (0.27).
Limitations of the study/implications: The results show the need to use local materials in the calibration process.
Conclusions: NIRS will make predictions of their chemical composition, since this is influenced by geographic origin of the sample and its botanical composition |
|---|